During the growth of p-doped semiconductor layers, an undesirable diffusion of the dopant, for example Mg or Zn, into semiconductor layers situated at a deeper level often occurs. Such diffusion of the dopant has been observed, for example in the case of InAlGaP layers which have been fabricated epitaxially in particular by means of MOVPE.
Particularly in the case of optoelectronic components, for example LEDs or semiconductor lasers, diffusion of the p-type dopant in the semiconductor chip is disadvantageous since it is known, for example, that the presence of Mg or Zn in the active layer of optoelectronic components may lead to nonradiative processes which reduce the light generating efficiency. Furthermore, an increased concentration of one of these dopants in the active layer of an optoelectronic component may bring about an accelerated ageing of the radiation-emitting component.
In order to reduce diffusion of the p-type dopant from the p-doped layer into the active layer, generally one or a plurality of undoped layers are inserted between the active layer and the p-doped layer. However, undoped layers of this type, with increasing thickness, have a disadvantageous effect on the optical and electronic properties of the semiconductor chip.
Furthermore, U.S. Pat. Nos. 6,233,266 and 7,042,013 disclose inserting a strained superlattice between a p-doped layer and an active layer as a diffusion barrier. The document JP 11068150 A furthermore proposes inserting a layer with lattice defects between the p-doped layer and the active layer as a diffusion barrier for the p-type dopant.